qavr-memory

What problem does it solve?

Memory systems often retrieve semantically similar information but fail to promote frequently useful memories. QAVR adds learned utility scoring to prioritize memories that led to successful outcomes, improving retrieval relevance over time.

Core Features & Use Cases

• Learned Q-Value scoring to rank memories by actual usefulness
• Cold-to-warm context adaptation: becomes active after sufficient interactions
• Integrates with a vector store to re-rank results for agents performing memory recall, planning, or debugging

Quick Start

Use qavr-memory to rank memories by integrating with your memory store and running a query like qavr_query(query_text, collection, n_results=5) to obtain prioritized results.